I will update this post as new information comes to hand
What has happened?
Recently the vaccine safety watch dogs in Europe noted reports of unusual types of blood clots in people vaccinated with the AstraZeneca (AZ) COVID-19 vaccine. This prompted investigations across many countries to ascertain what, why, and how. Also, in the last couple of days, the U.S. has reported unusual types of blood clots in people vaccinated with the Janssen/Johnson & Johnson COVID-19 vaccine.
Observations so far suggest that in younger AZ vaccinees, primarily women, there might be an unusual type of blood disorder that involves both low platelets (that can lead to bleeding) and blood clotting at the same time (thrombocytopenia and thrombosis). This disorder is appearing a week or two after vaccine administration and has been given several names so far including Thrombosis with Thrombocytopenia Syndrome (TTS). Importantly this syndrome also exists in association with other medications, infections, and sometimes with no known cause.
What are the risks?
While investigations are ongoing, the risk of developing this syndrome after receiving either the AZ or Janssen viral vector vaccines is miniscule. You are truly more likely to be struck by lightning in your lifetime. However, for those who do develop these blood clots the consequences can be dire with ¼ dying, and this is why the condition is being taken very seriously. The higher number of cases appearing in women may be because more women have been vaccinated.
Here are some comparisons
Why are we picking this up now?
We are picking this up now because of the sheer number of people being vaccinated at the same time (~820 million doses globally since December). Big numbers are really the only way very rare side effects can be detected. Even if these events occur in a trial there is no statistical power to tell if the event is by chance or not, they are not big enough. The fact that this has been identified is an indication that the safety monitoring systems are working pretty well. These are the surveillance systems that watch for adverse events after medicines start being used in the population. While some alerts turn out to be false alarms, others turn out to be real.
Were there any blood clots in the clinical trials?
There were thrombotic events in the trials, and here there is a perplexing pattern. Looking at thrombotic, thromboembolic, and neurovascular events in the AZ trials there were 7/12,282 in the vaccine group and 18/11,962 in the control group. Yes, there were over twice as many in the group that did not get the vaccine. And it gets weirder, in the Janssen, Pfizer, and Moderna vaccine trials it was the other way around.
In all these other vaccine trials there were numerically more clotting events in the vaccine group although these will be within the normal background rates and also not statistically significant, in other words taken in isolation there is no reason to think this is more than chance. Although, if you combine them then they probably become significant (I am not qualified to do the maths). HOWEVER, after hundreds of millions of doses of Pfizer and Moderna we do not see a signal. We should have seen one by now if they also had a problem. How can this all be upside down?
As the trials were all randomised, the vaccine and placebo recipients had equal chance of developing blood clots, unless there was a vaccine effect. The only difference between these groups of people was that almost all the COVID disease occurred in the placebo recipients and pretty much no vaccine recipients went to hospital with COVID. This all raises the possibility that there is another explanation for what we are seeing, or at least a more complex one. The fact that we are only seeing this with the viral vector vaccines suggests a vaccine class effect.
I think the clinical trail data may be a bit of a distraction. Trials fall way short of enough power to detect these rare events. We need big data.
How will we know if this condition is truly caused by the vaccines?
While we have seen reports of unusual blood clots, this does not mean they are definitely caused by the vaccine. There are three more steps that need to be completed.
- First, we need a really good definition of this condition such as laboratory results and the type of clots etc so we can look cases properly. See here for the evolving Brighton Collaboration case definition.
- Second, we need to compare what we are observing to what we expect and to do this we need to know how many cases of this condition occur normally in the background. For example, how many cases occurred in 2019? How many occurred in the last ten years? In whom? Once we gather this information then we can compare what we are seeing now with the normal patterns.
- Third, we need to conduct risk studies where we compare the risk of the condition in vaccinated people with the risk in unvaccinated people. There are a number of scientific methods to do this. If there is no difference then we might conclude that the vaccine is unlikely to be causing the problem, alternatively, if there is a greater risk in the vaccinated people then we might conclude that the vaccine is causing some of these events.
What do we do if the vaccine is to blame?
If the risk is as small as it currently appears to be, then we need to ask ourselves if it is worth using these vaccines and if so in whom. In a country with no COVID-19 there is no risk of blood clots from natural COVID-19 infection. If there is a vaccine that does not carry this risk (so far this risk is not associated with the Pfizer or Moderna vaccines) then it may be a better choice. However, in a setting where there is COVID-19, or no alternative vaccine available, then the AZ and Janssen vaccines will save many, many lives and also dampen the risk of dangerous variants developing. They are still very safe vaccines by all accounts. While these vaccines are halted because of a 1-4 in a million risk of blood clots (and <1 per million death) many thousands have lost their lives due to COVID-19 and the speed at which we might otherwise control this plague is slowed.